Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    INDEX
    Explanations

    end of sentence punctuation

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Perhaps
    1.19
     perhaps
    1.10
     wondrous
    1.10
     irony
    1.09
     truths
    1.05
     truth
    1.05
    perhaps
    1.00
    Perhaps
    1.00
     sorrow
    0.97
     wounds
    0.97
    POSITIVE LOGITS
    And
    0.80
    *[
    0.75
    I
    0.75
    Um
    0.72
    ['
    0.70
    receiving
    0.69
    即
    0.68
    ["
    0.68
    \&
    0.66
    H
    0.66
    Act Density 0.044%

    No Known Activations